OpenAI launched advertising inside ChatGPT in January 2026 after its CEO spent years publicly saying he hated the idea. The early numbers looked promising. The program hit $100 million in annualized revenue within six weeks, with over 600 advertisers joining. Then the context arrived.
OpenAI's own internal targets set $2.5 billion in ad revenue for 2026, rising to $100 billion by 2030. That 2030 figure would represent roughly half of Meta's current total annual ad revenue, a business Meta has been building since 2007. To get there, OpenAI would need to reach 2.75 billion weekly users by 2030. As of early 2026, the number stood at 900 million.
Emarketer, an independent research firm, looked at the same market and came to a starkly different conclusion. Its estimate puts total chatbot ad revenue across all platforms below $1 billion for 2026. For 2030, the firm projects $5.41 billion for the entire category. That is every chatbot combined, not just ChatGPT.
The gap between OpenAI's projections and independent estimates is not a matter of optimism versus pessimism. It reflects a real problem with how the ad platform works today. Advertisers report that the tools for measuring whether their spending is producing actual sales remain thin. Early in the rollout, OpenAI's reporting dashboard showed only two metrics: impressions and clicks. No conversion data, no detail on who saw the ad, no ability to confirm whether a sale actually happened downstream. Those tools have improved, but the platform is still well behind what marketers expect from Google or Meta, where advertisers can trace spending directly to revenue.
The format itself creates a measurement puzzle that does not exist in search advertising. When someone types a query into Google, sees an ad, clicks it, and buys something, the chain is trackable. When someone has a multi-turn conversation inside ChatGPT, sees a sponsored message, continues chatting, and then buys something three days later on a different device, that connection is much harder to prove. Agencies working on the platform have told industry press they cannot yet show clients clear business outcomes from ChatGPT ads.
This matters most because of what is coming next. OpenAI filed confidential paperwork with the SEC in June 2026 to prepare for a public listing, and is now leaning toward a 2027 debut. The company lost $38.5 billion in 2025 on $13 billion in revenue. It is burning cash at a scale that requires either a massive IPO, continued private fundraising, or a new revenue stream that actually delivers. Advertising was positioned as that stream.
The IPO math is already complicated. OpenAI's CEO has reportedly told advisers that listing below a $1 trillion valuation is a non-starter. The company's last private funding round valued it at $852 billion, and reaching $1 trillion requires investors to believe in future revenue that does not yet exist. A weaker-than-projected advertising business makes that case harder.
For any business operator thinking about whether to advertise on ChatGPT, the current state is worth understanding clearly. The platform is real, the audience is large, and the self-serve tools opened in May 2026. But the reporting is still immature compared to what you can get from Google or Meta, and independent analysts are not convinced the market will grow at anything close to the pace OpenAI needs. If you are curious, a small test budget makes sense. Committing serious spend before the measurement tools catch up does not.